I picked up djing last year. I had thought about it on and off for a while. What changed was watching friends grab an entry-level controller and play their first gig within a couple of months. Given the state of technology available, it’s become less technical and the barrier to entry became low enough for me and many others to hop over. That is true for a lot of fields with new AI tools.
Djing?
There are various descriptions of what being a DJ actually means. “90% troubleshooting and 10% playing songs” refers to the amalgamated equipment at most parties outside of dedicated music venues. “A glorified Spotify playlist” is the realisation that even very successful DJs follow a simple recipe when mixing tracks. Pick a next song that’s roughly in the key of your current track. Match the speed and align the beats, either by ear or by pressing the sync button. Then use the equalizer to swap the bass, mids and highs one by one. The transition should feel seamless, never breaking the hypnotic four beats per bar (I’m referring to techno here).
The build-ups and drops of modern electronic music are part of the tracks, even if some like to pretend otherwise by twisting dead knobs. It’s possible to accentuate certain parts by using filters. But the amount of filtering is sublinear to the quality of the mix. Less is more.
AI everywhere
The developments in generative AI have led to the first AI DJs. The software attempts the same workflow described above. The results are good enough and will soon transcend the skills of most bedroom DJs like myself. Djing now confronts the same question that many fields struggle to answer. Given the quality and proliferation of automated DJ software, where’s the human in the mix (excuse the pun)?
Advice heralds from an unlikely source: Jeff Bezos. He asks his managers to focus on the things that will never change. For Amazon, it means no customer will ever fault the company for having a larger selection, lower prices or faster delivery. These characteristics of the business will never change, and if in doubt, Amazon should work on those. There is a corresponding list of things for any activity or business.
For DJs, it is to play the right song at the right time. Lights, transitions, temperature or weather – none of it matters if it’s the right song for that moment. Everyone who has ever stood in ankle-deep mud or a sweaty basement knows the feeling. DJs read the crowd and notice when parts start to chat to each other or leave to grab a drink. Wedding DJs know if the girls start to dance, the guys will follow. To quote a groom at a recent wedding, exasperated by his DJ’s track selection: “he plays the right artists, but only their fifth most popular songs!”
Focus
If we accept that there is a right song for any point in time, it also means that DJs need a large library of tracks to select from. Spotify, Soundcloud, Tunebat and Bandcamp, to name a few, have made crate digging much easier. And yet in the end, it’s the DJ that needs to decide which of two tracks with popularity at 85 and energy at 68 is the right one. When I started out I thought that getting the transitions right and using filters is what makes or breaks a set. It couldn’t be further from the truth. A good DJ understands other people on an emotional level, emphasizing with the place and the moment in time. A good DJ is human in the social animal sense of the word. It’s our USP. The lesson here is the same as in any other business in the AI age: Focus on those aspects of your work that AI can't replace. Then select those that never change, like Bezos advises. And then double down. I’ll go back to my playlists now.